Actually, I find that a “web”
is a tangled mesh in which to entrap and is a term that infringes on other
trademarks and is a symbol that doesn’t inspire trust at all – said
the fly to the spider. I think repositories only hold value if the storage and
retrieval of information can be validated to be trusted, secure, reliable, and
honest in its representations. Since trust is an intangible value, I find it
to be an unobtainable goal in ecommerce activities based on concepts that are
not managed in accordance with good policies or even their own statements. IMHO
any system that is self-managed is quite frankly a disaster and provides me as
an informed consumer no level of trust or assurances proclaimed by any system
or solution said to be good for goodness sakes. Self managed is equivalent to
saying trust me because I say so.
Also, the wallet idea has come and gone
and a second showing will most likely result in the same…I also don’t
understand why it is referred to as a “centralized” repository and
then it is explained to be a “group” of repositories. Maybe it is
just me, but isn’t that a “distributed” model? Also, how do
you ensure the individual is always in control of their key? I think that
statement is simply not true, grandiose, and terrible misleading – and
not a basis of building any trust; as no system could ever ensure that the
action of all its individuals was consistent and the same. Therefore, the
International Web of Trust is seriously flawed by allowing such an assumption
to be the core value of its model.
I really think this whole message has been
completely self-promoting and quite a story about X.509 certificates but
vaguely disguised as a wallet, and I think the International Web of Trust is a
trap of old concepts, proclaiming protections that are perceived based on
biased assessments, and one of the most frightening concepts for ensuring trust
and identities that I have listened to in quite awhile.
This is exactly why such organizations as
IETF, ANSI, ISO, AICPA, NIST and Kantara are so much in need to prevent such
flawed solutions as this “web of trust” from being accepted by the electronic
commerce community.
(My corporate cell
phone number has been changed, if you could please update your records that
would be great.)
Sincerely,
Brian
Brian Dilley CISA / CIPP / CGEIT
President
GSA
Advantage! - Use our GSA Schedule 70 Today!
Office: (866) 465 - 6005
Fax: (443) 957 - 9005
Cell: (443) 955
- 9885
This electronic message contains
information from eValid8® that may be confidential,
proprietary or otherwise protected from disclosure and is
only intended for the recipient. If you should have received this transmission
in error, we do make mistakes when selecting addressees, please notify the
email originator at info@evalid8.com and
please delete that email message from your mailbox. eValid8®
wants to be safe on the Internet, and we honor clients and people’s
privacy. To read our Privacy Statement, click on this link, http://www.evalid8.com/contactus/privacystatement.html
.
From: community-bounces@kantarainitiative.org
[mailto:community-bounces@kantarainitiative.org] On Behalf Of Michael Duffy
Sent: Saturday, January 30, 2010
12:49 PM
To: Joni Brennan
Cc:
community@kantarainitiative.org
Subject: Re: [Kantara - Community]
Institutional Web of Trust
As requested, we have voided all company references
and provided a brief summary. Again, we apologize for the inappropriate
post.
Here is an overview of the Institutional Web of Trust
concept:
*************************************************************
Executive Summary: For the purpose of securing identity,
rather than creating a centralized directory or multiple directories of private
information, it will be far better to create a central repository containing a
collection of institutional decisions which will establish an Institutional Web of
Trust. In essence, there are a limited number of institutions worldwide
(measured in thousands) that truly matter when it comes to legitimizing
identity. Digital wallets on smart phones will enable the efficient
association of unique public/private keys to a specific legal identity (legal
name and legal address). If there is a non-unique association, an
inconsistency arises in the system. If the association is unique and
verified by one or more legitimate institutions an individual's identity is
secure (as long as the private key which he/she controls is secure).
This system secures identity, protects individual privacy and prevents the
establishment of monolithic government control. Under this system, the
user is always in control of his/her credentials.
This system makes a simplified and efficient federation process possible.
An institution could federate the identity
of it's users (or a subset of its users) simply by adding (or modifying) a
credential to each of it's user's digital wallet and creating an institutional
reference within The Institutional Web of Trust Repository.
This system provides the "Holy Grail" for single sign on.
*************************************************************
Digital credentials on NFC enabled smart phones will soon transform the world
of identity management. Within three years there will be
corporate and government deployments where all members of the organization are
issued NFC enabled smart phones for the purpose of identity management.
The basic question is, how can trust be established in the digital age?
If you and I have never met and I come to your website or place of business,
how can you be confident that I am who I say that I am? The Institutional
Web of Trust will resolve this basic question regarding the establishment of
trust.
A key component of the infrastructure will be an easy to use digital wallet where credentials can be
securely provisioned and transactions occur smoothly. This digital wallet will
be the cornerstone of NFC technologies on
mobile devices and provide the interface for identity, marketing and
financial services. Every aspect of digital life that depends on identity and
transactions will flow through the digital wallet.
This identity infrastructure will eliminate the possibility of identity theft
for all participants, protect consumers and financial institutions from
fraudulent transactions, greatly reduce cyber-crime and solve many of the
systemic problems of the current Public Key Infrastructure system, especially
the problems of certificate revocation lists (CRLs) and on-line status
checking.
The solution is simple, practical and transparent to the consumer. Consumer
acceptance will be rapid and widespread. The solution secures identity,
protects individual privacy and prevents the establishment of monolithic
government control. Under this system, the user is always in control of
his/her credentials.
The essence of the approach is very different from the "Big Brother"
approach recently announced by
Compared to a decentralized web of trust which creates a web of individuals
with, "the expectation that anyone receiving [a list of signatures] will
trust at least one or two of the signatures", we will create a system
where trusted
institutions legitimize individual identity. Additionally,
the Institutional
Web of Trust will have centralized controller processes that
rely greatly on self-management and automation resulting in great efficiencies.
Digital wallets on NFC enabled smart phones will enable users to secure their
private keys and control/present their digital credentials. Because a user's
identity will be authenticated by the processes of The Institutional Web of Trust
(not a trust authority) there is no need for a trust authority to issue and
vouch for public/private keys for individual users. It is only necessary that
the public key be registered and the private key be secured. Users can
self-issue their keys.
The
Institutional Web of Trust does not secure identity by,
"making personal data harder to steal". Rather, identity
is secured by self-managing logical inconsistencies within the system,
resolving identity conflicts and preventing fraudulent transactions.
As Bruce Schneier, author and security guru, pointed out, "Proposed
[identity theft] fixes tend to concentrate on the first issue--making personal
data harder to steal--whereas the real problem is the second [preventing
fraudulent transactions]. If we're ever going to manage the risks and effects
of electronic impersonation [identity theft], we
must concentrate on preventing and detecting fraudulent transactions."
[Solving Identity Theft]
In essence, there are a limited number of
institutions worldwide (measured in thousands) that truly matter
when it comes to legitimizing identity. Digital wallets on smart phones
will enable the efficient association of unique public/private keys to a
specific legal identity (legal name and legal address). If there is a
non-unique association, an inconsistency arises in the system. If the
association is unique and verified by one or more legitimate institutions an
individual's identity is secure (as long as the private key which he/she
controls is secure).
In the process of adding a credential to a user's digital wallet, the
provisioning institution (government agency, bank, university, etc.) will
calculate a secure hash value (numerical representation) of the credential
combined with information from the user's primary credential
(legal identity). This hash value will be encrypted with the user's
private key and then encrypted again with the provisioning institution's
private key; this encrypted hash value will then be stored in The Institutional Web
of Trust Repository representing an institutional validation of the user's identity.
This dual encryption establishes that the credential was associated with the
user during the provisioning process rather than simply asserting the
association by a reference from the repository. Also, There
is no need to store any specific information (account number, balance, etc.)
about user's account. The user is in complete control of the information
he/she presents and his/her privacy is maintained.
When a user presents a credential from his/her digital wallet a transaction ID
will be sent from the authenticating system to the user's digital wallet, be
encrypted with the user's private key and sent back to the authenticating
system. The user can be authenticated by decrypting the transaction ID with the
user's public key from The Institutional Web of Trust Repository.
The credential can be authenticated by calculating the hash value of the
credential and then decrypting the hash value stored in The Institutional Web
of Trust Repository with the institution's public key and the
user's public key.
In a variation of this process the provisioning institution does not store the
encrypted hash value in The Institutional Web of Trust Repository;
rather, the provisioning institution itself maintains a repository and a
reference to the repository is authenticated by an entry contained within The Institutional Web
of Trust Repository (through the institution's primary
credential). In this way an institution could federate the identity of it's users (or a subset of its users)
simply by adding (or modifying) a credential to each of it's user's digital
wallets and creating an institutional reference within The Institutional Web
of Trust Repository.
As part of the federation process, cooperating institutions will most likely
create standard authorization levels for various services and provision these
levels as part of a user's credential. For example, a coalition of
universities may have authorization levels for library services that will
enable users to access any library within the coalition; government
organizations may provision security levels within a user's credential that
enable inter-agency access to resources; etc.
There is significant debate regarding the effectiveness of biometrics in
identity management. When a user is not present (authenticating over a
network) there are fatal problems with biometric authentication. Most
significantly, "The main security problem with biometrics is the inability
to create a new secret. If you allow your fingerprint to be digitized and sent
across a network or scanned by a compromised scanner, it can be stolen. Then
someone has a digital copy of your fingerprint."
Even if a method of biometric identification proved to be completely reliable,
security issues would still remain. There would be opportunities to steal
someone's biometric signature and forge their identity credentials, especially
if there was a massive store of private personal data; one successful attack
could essentially render the entire system ineffective.
When a user is present bio-metric data can be an effective authenticator.
It will be possible to store bio-metric data
within a user's credential (not within a central repository) when
the credential is created by the provisioning institution. When a user
presents the credential, verifying the biometric data in the credential against
the individual in real time will provide enhanced security along with verifying
the encrypted transaction code against the user's public key and verifying the
encrypted hash code of the credential against the value stored in The Institutional Web
of Trust Repository.
While there are many types of biometric identifiers, one of the simplest and
most usable is a photograph of the human face verified by a human being.
Any credential in a user's digital wallet that includes a photograph (driver's
license, passport, bank debit card, etc.) will be highly reliable when a user
presents the credential in person.
Why would a major institution (bank, university, corporation, government
agency, etc.) utilize The Institutional Web of Trust Repository
instead of its own internal system? When there is no need for an external
third party to rely on a user's credential an institution may very well utilize
its own internal repository. In this same case, smaller institutions, for
reasons of convenience and cost, might still utilize The Institutional Web
of Trust Repository.
Whenever a third party (a party other than the provisioning institution) must
relay on a user's credential, the key services The Institutional Web of Trust Repository
provides are assurance that the user is unique and trustworthy, assurance that
the provisioning institution is unique and trustworthy and assurance that the
credential is trustworthy. Also, The Institutional Web of Trust Repository
creates a "data synergy effect"
which establishes an Institutional Web of Trust (when multiple
institutions validate a unique user's identity the identity becomes more secure
and trustworthy).
If a unique user has digital credentials for a state driver's license, a
passport, a bank debit card, a university ID, insurance cards, credit cards,
etc., all independently validated by trustworthy institutions, that user's
identity is secure and highly trustworthy. Similar to credit
ratings, both individuals and institutions will have "trust ratings" within The Institutional Web
of Trust Repository. A centralized notification service
will also be provided when credentials are lost or stolen.
The uniqueness test for legal identities within The Institutional Web of Trust Repository
helps to secure identity and prevent identity theft. If there is a
non-unique association, an inconsistency arises in the system. Also, easy
access for online status checking establishes the currency of a user's
credentials in case the user's digital wallet is lost or stolen.
Additionally, the system provides the "Holy Grail" for single
sign on. All computers will soon have an interface (USB
plugin or internal card) that will enable NFC interactions with mobile devices.
The digital wallet on a user's cell phone will be provisioned with credentials
containing specified authorizations for different systems and services. Rather
than logging into a directory or utilizing a complex federated identity
process, a user will log onto his/her cell phone with a PIN and/or a voice
authentication signature. The user (or the authenticating system) will then
select the appropriate credential for the specified system or service with no
need to enter another user name or password (the user's private key will be
used to encrypt a transaction ID). This approach also solves the "Keys to the Kingdom" problem
where a single sign on to a directory service opens access to all the user's
systems and services.
Additionally, the system will enable a process of mutual authentication that will prevent phishing scams.
The user's credential and the institution's credential could both contain a
list of valid URLs which could be matched during the sign on process.
Existing providers of identity management services should not see The Institutional Web
of Trust as a competitor; rather, they should see it as an
infrastructure service (similar to the electric power grid that has hundreds of
energy providers).
This identity infrastructure will be created with government resources and be
managed to a great extent as a public trust.
Best regards,
Michael Duffy
CEO / CTO ~ The Trust Nexus
http://www.thetrustnexus.com
Joni Brennan wrote:
Thanks Mike,
It seems like you have a product and you're telling us about it. We don't
allow these types of messages here. So you have a few choices of how to
interact.
1 - Parse your email down to a very brief idea and some questions that the
group could respond to. (This mode is more aligned with the true nature
of the list.) Again - as it reads now it's a product advertisement which
is prohibited.
2 - Browse our groups list from the homepage here http://kantarainitiative.org.
You may find a group you could join and then share your idea there.
3 - Start your own Work or Discussion Group to discuss your identity based
solutions ideas. If you're interested in this path please ping me
directly so we can discuss and I can learn more about your goals.
>From there I could help you to determine if this is appropriate work
material for Kantara or not.
Cheers - Joni
On Fri, Jan 29, 2010 at 10:04 AM, Michael Duffy <thetrustnexus@austin.rr.com>
wrote:
My apologies.
A member of your group actually suggested we post to the community list.
Our goal was not to promote the company but to discuss the ideas related to an
Institutional Web of Trust with the leading experts in the field of identity.
Is there another existing list that would better suit this purpose?
If not, perhaps Kantara could create such a list (e.g., ideas@kantarainitiative.org).
Mike
Roger Sullivan wrote:
Let me echo Brett's sentiments. None of Kantara's lists are to be used for promotion of oneself or one's company. Please refrain from doing that. Roger SullivanPresident, Kantara Initiative -----Original Message-----From: Brett McDowell [mailto:email@brettmcdowell.com] Sent: Friday, January 29, 2010 11:04 AMTo: Thomas HardjonoCc: community@kantarainitiative.orgSubject: Re: [Kantara - Community] Institutional Web of Trust That is certainly not its intent and if it becomes used in that way Ifear we will see mass un-subscription which will undermine our abilityto communicate with each other. So I would ask that all subscribersrefrain from using this list for any form of advertising. Thank you, || Brett McDowell, Executive Director, Kantara Initiative On Fri, Jan 29, 2010 at 10:59 AM, Thomas Hardjono<standards@hardjono.net> wrote: